Telecoms.com periodically invites expert third parties to share their views on the industry’s most pressing issues. In this piece Yuval Stein, AVP Technologies at TEOCO, examines best practice for autonomous networks.
As 5G proliferates across a number of industry verticals, the vision of a fully connected world arrives with new technical challenges as well as myriad opportunities. As network operators strive to meet rising customer expectations, deliver increasingly digitized and diversified services, all the while trying to lower OPEX, many are contemplating the ways to more easily manage and oversee their networks.
One thing is clear, operators will struggle to provide high-quality services and maintain their competitive advantage if they continue to rely on the same manual, labour-intensive methods of operating and maintaining their networks as they strive to deliver the latest 5G services. This is where autonomous networks might offer a solution.
An autonomous network is a network able to configure, monitor and maintain itself independently and can offer self-healing and self-optimization capabilities. These networks can run based on the intent determined by an operator’s desired business outcomes and are able to adapt to their environment and learn from data by leveraging AI and machine learning. The autonomous network is adaptive, agile, and programmable, and enables the operation of the network to be aligned with the goals of the business and runs with minimal to no human intervention. As a result of these characteristics, the network becomes a strategic tool for operators as long-term planning becomes an option.
New network management: autonomy is key
Ever since mobile networks were first deployed over 30 years ago, their infrastructure has required the continuous oversight of thousands of engineers, each observing specific parameters and then intervening to troubleshoot issues as they occur. This way of managing the network comes with two obvious disadvantages – not only is this type of manual oversight extremely resource intensive, but it is also invariably error-prone in the face of human fallibility.
Now with the rollout of 5G networks, this approach demands a rethink. With highly dynamic and multi-faceted 5G networks supporting an ever-expanding array of connected devices and machines, it is almost impossible for even extensive engineering teams to be across everything taking place on the network at any given moment — a commercial network generates billions of transactions and vast volumes of traffic on a daily basis. The absence of a complete and comprehensive awareness of the network and its underlying infrastructure presents a serious risk for operators given that changes or updates in one domain may have unintended and unforeseen consequences somewhere else.
5G and its associated technology ecosystems (IoT, AI, edge compute and robotics) are the substrate on which enterprise business and operational models can be transformed. For instance, functionalities such as network slicing will allow operators to deliver services customized to support key industry verticals. In light of these advancements, the traditional ways of managing networks are no longer viable. Automation and network autonomy are essential to manage growing complexity, particularly as the network management and control required to overcome this complexity transcend the capabilities of humans and also straightforward pre-defined automated processes.
The drive for autonomy in network and operations
Meaningful automation across an operator’s business ecosystem requires the buy-in of various stakeholders across the organisation, and, in particular, the commitment of the executive leadership team. In fact, the consensus amongst many operators is that autonomous networks are crucial to their long-term strategic success: they will be able to benefit from a strong improvement in efficiency, and the innovative services that autonomous networks enable will present new revenue opportunities in new markets. As a result, many are now seeking policies and the practical means to keep the goal of autonomous networks on track.
This starts with scaling up their ambitions and taking a more macro view of the network. Many of the capabilities operators need to lead in the digital economy rely on seamless automation. Operators have traditionally taken a pragmatic, piecemeal approach to automating individual domains, but the absence of an integrated plan for building autonomous networks is holding them back.
Formulating such a plan is not always easy when the vast majority of operators still have to contend with maintaining legacy systems – these remain the primary challenge for operators looking to automate their networks and operations. The main processes including design, planning, fulfilment, assurance, and maintenance are highly fragmented in the current mode of operations, and often require manual interventions and inter-departmental handovers. Generally, this has been due to operators implementing various siloed network technologies, operational support systems (OSS) and other ad-hoc tools, resulting in non-integrated and oftentimes disjointed network operations.
The rise of autonomous networks necessitates a fundamental break with this model. Operators have to simplify and streamline their entire operations and join up all of the various processes end-to-end through domain and cross-domain closed-loop mechanisms with software as well as artificial intelligence (AI) and Machine Learning systems. Only then will they be able to actualize full lifecycle automation.
Autonomous networks: the next phase of telco evolution
Today, the operators making good progress towards achieving an autonomous network largely rely on a hybrid model – this in practice means a combination of human and automated oversight. In fact, some are now looking to deploy fully autonomous systems in parts of the network where actions are relatively simple to define. A significant step on the path to complete autonomy is the introduction of self-learning capabilities in the network and, with it, adaptation. And with the technology continuing to advance, such systems will be able to take on greater responsibility, resulting in increased reliability and improved quality of service (QoS).
In the coming years, autonomous networks will be foundational to achieving operational agility which in turn will lower OPEX. This is significant as success in the 5G era rests on an operator’s ability to launch and operationalize new digital services for a fraction of their operating cost. At the same time, a network that is self-organized and self-optimized will be critical to ensuring the service continuity that enterprises demand. As operators leverage the power of autonomous networks, they will be in a more favorable position to improve operational efficiency and support business innovation through a more agile approach.
Yuval is currently the AVP Technologies at TEOCO focusing on the research and design of OSS service assurance solutions for the evolving telecommunication technology, mainly Network Function Virtualisation. He has previously held senior product management and technology positions in the telecommunication management software industry for the past 15 years.
Get the latest news straight to your inbox. Register for the Telecoms.com newsletter here.